Companies are investing in AI, launching pilots, and training their people. Yet many are still struggling to turn all that activity into lasting business value. Why do so many AI transformations stall? One important—albeit sometimes hard-to-admit—reason for such stalls is unproductive leadership habits. Let’s talk about it.

Part of the challenge is that the technology keeps moving while the way people work stays much the same. Getting value from AI requires an adaptable culture: one where people are willing and able to change how they work as better approaches emerge.

In an adaptable culture, people understand the outcome they’re working toward and have room to figure out a better way to get there. They test ideas, talk openly about what happened, and put what they learn to use. When something doesn’t work, they can say so. When a familiar process stops making sense, they can question it. Useful discoveries get shared across teams and become part of how the organization operates. Over time, these behaviors of adaptability become the way work gets done.

One of the key drivers to building an adaptable culture is leadership. No surprise there. The tricky part is that some perfectly understandable leadership habits can reinforce the very behaviors they’re hoping to change. Here are three to watch for, and what leaders can do to successfully lead through AI transformation.

Three ways leaders get in the way

Only the Good News. “Experiment, but make sure it works.” Leaders encourage people to try new things with AI, then ask only for polished success stories and predictable results. Questions get brushed aside. Failed attempts disappear from the discussion. People quickly figure out what is welcome on the topic and what is too politically risky to share. The team looks confident, even though they feel shaky at best, and the conversations that would better direct AI efforts just aren’t happening.

My Way or the Highway. “Find a better approach, but follow these steps.” Leaders define the goal, prescribe the solution, and require approval for every adjustment. When a team discovers something useful, changing direction becomes a sensitive negotiation. Eventually, people stop offering alternatives and focus on following instructions. The organization gets compliance when what it really needs is people exercising judgment.

On Your Own Time. “Keep learning, but don’t let anything slip.” Leaders ask people to explore AI, build skills, and rethink ways of working while leaving every existing commitment in place. Come Monday morning, the calendar is full, the deadlines haven’t moved and people are pulled in more directions than ever. A few enthusiastic employees squeeze it in. A few people squeeze in an experiment. By the next Monday, the team is back to its usual routine.

Three ways leaders can build an adaptable culture

Pave a Safe Path. An adaptable culture depends on people being willing to tell leaders what they’re seeing, especially when an AI experiment falls short or raises uncomfortable questions about how people’s roles might change. Leaders set the tone in those moments. When they’re open about their own missteps and curious about what others are finding, people are more likely to surface problems while there’s still time to do something about them.

Take action. At your next leadership meeting, share an AI experiment of your own that fell short and what you changed. Ask each team to share one of their own failures and what they learned. Celebrate a useful lesson, not just a successful result.

Create Clarity on Purpose. Teams can adapt more easily when they know what they’re trying to improve and which decisions they can make along the way. A clear business outcome gives them a way to judge whether AI is helping. Clear boundaries for responsible use give them room to try different approaches and change course as they learn.

Take action. Pick one business process where AI could help. Agree with the process owner on a measurable target, establish data/security guardrails, and authorize the team to test, learn, and implement changes without requesting approval for every iteration.

Create Capacity for Change. Learning to work differently takes time. People need room to practice, compare approaches, and work out how to make a better approach part of the job. That room comes from decisions leaders make about what work takes priority and what can wait.

Take action. Select one AI experimentation team, dedicate four hours per person per week for 30 days, and explicitly remove or defer other work to protect that time.

Start with a Conversation

Improving the adaptability of your culture starts with everyday conversations and experiences, which you as a leader have a large role in shaping. To get the conversation started, sit down with your team and ask:

• Pave a Safe Path. What’s one concern about AI we haven’t talked about openly?
• Create Clarity on Purpose. What outcome are we trying to improve with AI, and what’s still unclear about how much freedom you have to try a different approach?
• Create Capacity for Change. What should we pause this month to give you time to test a better way of working?

Want to go deeper? Take our AI Readiness Assessment to explore where your organization is ready to move forward and what may be getting in the way of turning AI into business value.